datalog: hash-set membership for facts (Phase 5b perf)

db gains a parallel :facts-keys {<rel>: {<tuple-string>: true}}
index alongside :facts. dl-tuple-key derives a stable string via
(str lit) — (p 30) and (p 30.0) collide correctly because SX
prints them identically. dl-add-fact! membership is now O(1)
instead of O(n) list scan; insert sequences for relations sized
N drop from O(N²) to O(N).

Wall clock on chain-7 saturation halves (~12s → ~6s); chain-15
roughly halves (~50s → ~25s) under shared CPU. Larger chains
still slow due to body-join overhead in dl-find-bindings —
Blocker entry refreshed with proposed follow-ups.

dl-retract! keeps both indices consistent: kept-keys is rebuilt
during the EDB filter, IDB wipes clear both lists and key dicts.

lib/datalog/api.sx

@@ -111,21 +111,29 @@
           (has-key? (get db :facts) rel-key)
           (let
             ((existing (get (get db :facts) rel-key))
-              (kept (list)))
+              (kept (list))
+              (kept-keys {}))
             (do
               (for-each
                 (fn
                   (t)
                   (when
                     (not (dl-tuple-equal? t lit))
-                    (append! kept t)))
+                    (do
+                      (append! kept t)
+                      (dict-set! kept-keys (dl-tuple-key t) true))))
                 existing)
-              (dict-set! (get db :facts) rel-key kept))))
+              (dict-set! (get db :facts) rel-key kept)
+              (dict-set! (get db :facts-keys) rel-key kept-keys))))
         ;; Wipe all relations that have a rule (these are IDB) so the
         ;; saturator regenerates them from the surviving EDB.
         (let ((rule-heads (dl-rule-head-rels db)))
           (for-each
-            (fn (k) (dict-set! (get db :facts) k (list)))
+            (fn
+              (k)
+              (do
+                (dict-set! (get db :facts) k (list))
+                (dict-set! (get db :facts-keys) k {})))
             rule-heads))
         (dl-saturate! db)
         db))))

lib/datalog/db.sx

@@ -12,7 +12,7 @@
 ;; lib/datalog/builtins.sx) swaps in the real `dl-rule-check-safety`,
 ;; which is order-aware and understands built-in predicates.
 
-(define dl-make-db (fn () {:facts {} :rules (list)}))
+(define dl-make-db (fn () {:facts {} :facts-keys {} :rules (list)}))
 
 (define
   dl-rel-name
@@ -97,13 +97,19 @@
   (fn
     (db rel-key)
     (let
-      ((facts (get db :facts)))
+      ((facts (get db :facts))
+        (fk (get db :facts-keys)))
       (do
         (when
           (not (has-key? facts rel-key))
           (dict-set! facts rel-key (list)))
+        (when
+          (not (has-key? fk rel-key))
+          (dict-set! fk rel-key {}))
         (get facts rel-key)))))
 
+(define dl-tuple-key (fn (lit) (str lit)))
+
 (define
   dl-rel-tuples
   (fn
@@ -125,10 +131,16 @@
         (let
           ((rel-key (dl-rel-name lit)))
           (let
-            ((tuples (dl-ensure-rel! db rel-key)))
+            ((tuples (dl-ensure-rel! db rel-key))
+              (key-dict (get (get db :facts-keys) rel-key))
+              (tk (dl-tuple-key lit)))
             (cond
-              ((dl-tuple-member? lit tuples) false)
-              (else (do (append! tuples lit) true)))))))))
+              ((has-key? key-dict tk) false)
+              (else
+                (do
+                  (dict-set! key-dict tk true)
+                  (append! tuples lit)
+                  true)))))))))
 
 ;; The full safety check lives in builtins.sx (it has to know which
 ;; predicates are built-ins). dl-add-rule! calls it via forward

plans/datalog-on-sx.md

@@ -246,17 +246,29 @@ large graphs.
 
 ## Blockers
 
-- **Hash-set membership for relations.** `dl-tuple-member?` uses a linear
-  list scan; insert is O(n) and saturating chain-N pushes O(n²) → O(n³)
-  total. Under bundled conformance (CPU shared with other loop agents)
-  even chain-15 hits multi-minute wall-clock. Tests scoped to chain-5
-  for now. Fix: maintain a `{tuple-key → true}` dict per relation
-  alongside the list; key tuples by their serialized form.
+- **Saturation perf on long chains.** Resolved one bottleneck (hash-set
+  membership in `dl-add-fact!`) but `dl-saturate!` still spends
+  significant time per iteration on rule body joins — chain-15 takes
+  ~25s real / 3s user under contention even after the membership fix.
+  Two follow-ups to consider: (a) avoid `(rest lits)` in
+  `dl-find-bindings`/`dl-fbs-aux` (uses indexed iteration like the
+  membership fix), (b) memoize the per-rule body shape so `(len lits)`
+  and accessor calls don't re-walk the list each step.
 
 ## Progress log
 
 _Newest first._
 
+- 2026-05-08 — Phase 5b perf: hash-set membership in `dl-add-fact!`.
+  db gains a parallel `:facts-keys {<rel>: {<tuple-string>: true}}`
+  index alongside `:facts`. `dl-tuple-key` derives a stable string
+  key via `(str lit)` — `(p 30)` and `(p 30.0)` collide correctly
+  because SX prints them identically. Insertion is O(1) instead of
+  O(n). chain-7 saturation drops from ~12s to ~6s; chain-15 from
+  ~50s to ~25s under shared CPU. Larger chains are still slow due
+  to body-join overhead in dl-find-bindings (Blocker updated).
+  `dl-retract!` updated to keep both indices consistent. 143/143.
+
 - 2026-05-08 — Phase 9 done. New `lib/datalog/api.sx` exposes a
   parser-free embedding: `dl-program-data facts rules` accepts SX
   data lists, with rules in either dict form or list form using